• No results found

The most effective way to carry out a fraud brainstorming session

N/A
N/A
Protected

Academic year: 2021

Share "The most effective way to carry out a fraud brainstorming session"

Copied!
33
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

The most effective way to carry out a fraud brainstorming session

Jim Bijvoet 10194215 30-06-2014

Bsc Accountancy & Control University of Amsterdam Supervisor: J.J. Schipper RA

(2)

Abstract

For years, fraud has been a problem for firms all over the world. In order to reduce fraud, SAS No. 99 was issued in 2002. This standard requires all audit teams to conduct a fraud brainstorming session during a financial statement audit. A fraud brainstorming session is a meeting of the audit team in which participants generate ideas about possible fraud in an organisation. There are multiple ways in which a session can be carried out. However, SAS No. 99 does not state how this should be done. Therefore, the research question of this thesis is: What is the most effective way to carry out a fraud brainstorming session? In order to answer this question, a literature review was performed. An analysis was made in which the three choices that the auditor has to make were examined. For every decision, there are two options. These options were compared in order to find the most effective methods to carry out a fraud brainstorming session. It was found that a fraud brainstorming session should take place either in a nominal or group session, in which electronic technology is used and audit team members are accountable for the ideas they propose about possible fraud. This finding contributes to the existing literature about fraud brainstorming sessions and the various ways in which a session can be carried out. Also, auditors who have to lead a fraud brainstorming session can apply this result in real-world practice.

Samenvatting

Al jarenlang is fraude voor bedrijven over de hele wereld een probleem. Om fraude te verminderen is SAS No. 99 geïntroduceerd in 2002. Deze standaard verplicht alle accountantsteams om een fraude brainstormsessie uit te voeren tijdens een jaarrekeningcontrole. Een fraude brainstormsessie is een bijeenkomst van het accountantsteam om ideeën te genereren over eventuele fraude in een bedrijf. Er zijn meerdere manieren waarop een sessie kan worden uitgevoerd. SAS No. 99 geeft echter niet aan hoe dit moet worden gedaan. De onderzoeksvraag van deze scriptie is daarom: Wat is de meest effectieve manier om een fraude brainstormsessie uit te voeren? Om deze vraag te beantwoorden werd een literatuuronderzoek uitgevoerd. Er werd een analyse verricht waarin de drie keuzes werden onderzocht die de accountant moet maken. Voor elke beslissing zijn er twee opties. Deze opties werden vergeleken om erachter te komen wat de meest effectieve methoden zijn om een fraude brainstormsessie uit te voeren. Het resultaat was dat een fraude brainstormsessie moet plaatsvinden in een nominale of groepssessie, waarin elektronische technologie wordt gebruikt en

accountantsteamleden verantwoordelijk zijn voor de ideeën die ze voorstellen over mogelijke fraude. Deze bevinding draagt bij aan de bestaande literatuur over fraude brainstormsessies en de

verschillende manieren waarop een sessie kan worden uitgevoerd. Ook kunnen accountants die een fraude brainstormsessie moeten leiden dit resultaat in de realiteit toepassen.

(3)

Table of contents

1. Introduction 4

2. Background information 5

2.1 Fraud 5

2.2 Statement on Auditing Standards No. 99 6

2.3 Fraud brainstorming sessions 7

2.4 Group and nominal fraud brainstorming sessions 8 2.5 Face-to-face and electronic fraud brainstorming sessions 9 2.6 Accountable and anonymous fraud brainstorming sessions 10

3. Analysis 11

3.1 Group and nominal fraud brainstorming sessions compared 12 3.1.1 Comparison in terms of quality of ideas 12 3.1.2 Comparison in terms of quantity of ideas 14 3.2 Face-to-face and electronic fraud brainstorming sessions compared 16 3.2.1 Comparison in terms of quality of ideas 17 3.2.2 Comparison in terms of quantity of ideas 19 3.3 Accountable and anonymous fraud brainstorming sessions compared 21 3.3.1 Comparison in terms of quality of ideas 22 3.3.2 Comparison in terms of quantity of ideas 24

4. Discussion 26

5. Conclusion 28

(4)

1. Introduction

For years, fraud has been a problem for firms all over the world. The Association of Certified Fraud Examiners (2014, p. 4) states that globally total fraud amounts to $3.7 trillion on an annual basis and that the typical organization loses 5% of its revenues to fraud. In order to reduce fraud, several measures have been taken over the years. One of these measures is SAS No. 99, an auditing standard which was issued in 2002. This standard requires all audit teams to conduct a fraud brainstorming session during a financial statement audit. A fraud brainstorming session is a meeting of the audit team in which participants generate ideas about possible fraud in an organisation. There are multiple ways in which a session can be carried out. However, SAS No. 99 does not state how this should be done. Therefore, the research question of this thesis is: What is the most effective way to carry out a fraud brainstorming session?

In this paper, a literature review will be performed in order to answer the research question. First, some background information will be provided about the different ways in which a fraud brainstorming session can take place. After that, there will be an analysis which will show the three different choices that the auditor has to make when he has to organise a session. For every decision, there are two options which can be chosen and these options will be compared. First, there will be a comparison of nominal and group fraud brainstorming sessions. After that, electronic and face-to-face fraud brainstorming sessions will be compared. The last part of the analysis will show whether accountable or anonymous fraud brainstorming sessions are preferred. Fraud brainstorming sessions are effective when many ideas about possible fraud have been produced and when these ideas are of high quality. Therefore, all three comparisons are split into two components: quality and quantity of ideas. After the analysis, a discussion will take place in which the results of the analysis will be evaluated in order to show which methods should be used to make a fraud

brainstorming session most effective.

The analysis provides clear results about the different decisions that the auditor has to make. In the first comparison, it is shown that nominal fraud brainstorming sessions generate most ideas, while group fraud brainstorming sessions produce ideas of higher quality. As a result, none of these methods is considered to be more effective than the other. The result of the second part of the analysis is that electronic fraud brainstorming sessions outperform face-to-face fraud brainstorming sessions in both quality and quantity of ideas. Therefore, electronic fraud brainstorming sessions are the most effective alternative. The final part of the analysis shows that accountable fraud brainstorming sessions are preferred above anonymous fraud brainstorming sessions. When audit team members are accountable

(5)

for their idea production, they generate more ideas and these are of higher quality than ideas from anonymous participants.

The results of the analysis clearly show the answer to the research question: a fraud brainstorming session should take place either in a nominal or group session, electronic technology must be used and audit team members should be accountable for the ideas they propose about possible fraud. This is the most effective way of carrying out a fraud

brainstorming session. By presenting this result, this thesis makes a contribution to the existing literature about fraud brainstorming sessions and the different ways of conducting a session. In addition, the findings of this paper will be useful for auditors who have to lead a fraud brainstorming session. They have to make several decisions about how a session will be carried out and the results of this thesis show what the best choices are.

In the next part of this thesis, some background information about important concepts will be provided and the different fraud brainstorming techniques will be explained. The third section will contain the analysis in which the different fraud brainstorming methods will be compared in terms of idea quality and quantity. In the fourth part of this paper, a discussion will take place about the findings from the analysis in order to show which methods are most effective. Finally, the last section will present the conclusion of this thesis.

2. Background information

In this section some background information will be provided about important concepts which are essential for answering the research question. First, fraud will be described. This will be followed by an explanation of a measure against fraud, SAS No. 99. After that, one of the requirements of SAS No. 99, fraud brainstorming sessions, will be described. The last three subsections will discuss the various ways in which fraud brainstorming sessions can be carried out: group or nominal fraud brainstorming, face-to-face or electronic fraud brainstorming, and accountable or anonymous fraud brainstorming.

2.1 Fraud

According to Maddox (2000, p. 19), fraud can be defined as “an intentional act that results in a material misstatement in financial statements that are the subject of an audit”. There are two types of fraud: fraudulent financial reporting and misappropriation of assets. Fraudulent financial reporting means that financial statements are intentionally misstated in order to mislead the users of those financial statements (Debreceny & Gray, 2010, p. 161). For instance, fictitious revenues may be recorded or losses may not be shown on the income

(6)

statement in order to create a better image of a firm’s financial situation. Misappropriation of assets occurs when employees steal assets from the firm in order to benefit from this personally (Maddox, 2000, p. 19). Examples of this kind of fraud include stealing cash or theft of inventory for personal use.

Morales, Gendron and Guénin-Paracini (2014, p. 170) state that there are three necessary elements which have to be present in order for fraud to occur. These combined elements are generally known as the fraud triangle. First, there has to be pressure for the perpetrator to commit the fraud (Dellaportas, 2013, pp. 30-31). The person has, for example, personal financial problems and desperately needs money in order to pay off his debt. The second element which is required is an opportunity (Dellaportas, 2013, pp. 31-32). The perpetrator may have found a way of committing a fraud and being able to cover his tracks. This opportunity might arise when internal controls are inefficient or when there is no clear separation of duties. The final element is a rationalization (Dellaportas, 2013, p. 32). This means that the person does not feel guilty about his actions and that he can think of a reason to make the fraud acceptable.

The Association of Certified Fraud Examiners (2014, p. 4) reports that globally total fraud amounts to $3.7 trillion on an annual basis. They state that the typical organization loses 5% of its revenues to fraud each year. In order to reduce fraud, several measures have been taken over the years. One of these is SAS No. 99, which will be explained in the next subsection. Despite these measures, tips are still the most common fraud detection method, adding up to discover over 40% of all detected fraud (The Association of Certified Fraud Examiners, 2014, p. 4).

2.2 Statement on Auditing Standards No. 99

In November 2002, the Auditing Standards Board issued a new auditing standard, which is aimed at detecting fraud. This standard is called Statement on Auditing Standards No. 99, Consideration of Fraud in a Financial Statement Audit, which is commonly abbreviated to SAS No. 99 (Marczewski & Akers, 2005, p. 38). The standard superseded SAS No. 82, which was issued in 1997 (McConnell & Banks, 2003, p. 27). Compared to the previous standard, the responsibility of the auditor remains the same: the auditor has to plan and perform an audit in order to obtain reasonable assurance about whether financial statements are free of material misstatements (Casabona & Grego, 2003, p. 17). However, SAS No. 99 introduces additional procedures and requirements to assist auditors in performing an audit.

Casabona and Grego (2003, pp.17-18) provide an overview of the different components of SAS No. 99. The standard stresses the importance of professional

(7)

should he assume complete honesty (Maddox, 2000, p. 20). SAS No. 99 also requires the audit team to carry out a fraud brainstorming session during a financial statement audit. The purpose of this brainstorming session is to generate ideas about how and where fraud might be occurring and how to detect it (Brazel, Carpenter & Jenkins, 2010, p. 1274). This concept will be explained in the next subsection. Furthermore, additional procedures are required to obtain information which is needed to identify fraud risks. This includes inquiries of

management and employees, and analytical procedures (Ramos, 2003, p. 30). In addition, analytical reviews of revenue during audit planning are obligatory and the standard also statesthat revenue recognition should be considered as a possible fraud risk. Finally, SAS No. 99 requires procedures to understand how controls can be overridden by management and procedures to understand the process for recording journal entries (Casabona & Grego, 2003, p. 17).

2.3 Fraud brainstorming sessions

As was mentioned in the previous subsection, SAS No. 99 requires the audit team to carry out a fraud brainstorming session during a financial statement audit. Lynch, Murthy and Engle (2009, p. 1210) describe a fraud brainstorming session as a meeting in which the audit team comes together to think of ideas on how and where fraud might be occurring in the firm which is being audited. The audit team also has to think about how management could possibly conceal the fraud and they also have to come up with ideas on how to detect the fraud (Hoffman & Zimbelman, 2009, p. 812). According to Carpenter, Reimers and Fretwell (2011, p. 212), the objective of the session is to generate many, high-quality ideas about potential fraud. This means that a fraud brainstorming session is effective when many ideas about possible fraud have been generated and when generated ideas are of high quality. Another aim of the fraud brainstorming sessions is that the importance of a questioning mind and professional scepticism is emphasised to all members of the audit team (Ramos, 2003, p. 29). This should improve the performance of all audit procedures and will therefore result in a more successful audit.

While the fraud brainstorming session is a mandatory element of a financial statement audit, SAS No. 99 does not provide clear instructions on when and how a session should be carried out. Ramos (2003, p. 29) reports that the brainstorming session does not necessarily have to take place during the planning phase of the audit process. Because data is gathered continuously throughout the audit, it can be carried out at any specific time that the auditor wishes to choose. Bellovary and Johnstone (2007, p. 1) state that the only requirement is that the discussion should involve key members of the audit team and that multiple factors

(8)

will influence the extent of the discussion and how it should occur. This allows auditors to choose their preferred way of carrying out a fraud brainstorming session.

There are multiple ways in which a fraud brainstorming session can be carried out. Lynch (2006, pp. 69-70) describes three areas in which the auditor has the ability to choose different brainstorming approaches. First, the auditor can choose between group

brainstorming or nominal brainstorming. Group brainstorming involves a meeting of the entire audit team in which ideas about possible fraud are generated jointly, whereas nominal

brainstorming lets group members generate ideas individually, after which the ideas will be shared in a group session (Lynch, 2006, p. 69). Another consideration that should be made is whether or not to conduct the brainstorming session electronically. Electronic

brainstorming involves the use of technology, such as a group support system, to

communicate ideas to other audit team members (Lynch, 2006, p. 69). The final decision which has to be made about the fraud brainstorming session is the possible use of

anonymity. When using anonymous brainstorming, group members can submit ideas about possible frauds without being accountable for these ideas (Lynch, 2006, pp. 69-70). These three different topics on which the auditor has to make a decision will be explained further in the following subsections.

2.4 Group and nominal fraud brainstorming sessions

In the previous subsection it was mentioned that the auditor is able to make several choices when it comes to conducting a fraud brainstorming session. One of these choices is the decision between a group fraud brainstorming session and a nominal fraud brainstorming session (Lynch, 2006, p. 69).

A group fraud brainstorming session is the most common brainstorming approach. It is described by Lynch (2006, p. 69) as an open, interactive discussion in which audit team members jointly generate ideas about potential fraud in a firm. The discussion is unstructured and team members can share ideas in a free-flow-manner (Bean, 2008, 24). In a group brainstorming session only one person can speak at a time. This may cause other group members to forget or suppress their ideas when they are not allowed to talk (Hunton & Gold, 2010, 915). According to Bean (2008, p. 24), during this type of brainstorming there is usually no equal participation between the audit team members. This type of session commonly ends up being dominated by one or only a few participating members.

Another fraud brainstorming approach is a nominal fraud brainstorming session. Lynch (2006, p. 69) states that in a nominal fraud brainstorming session all individual members of the audit team brainstorm on their own to generate ideas about possible fraud. After this, a group session takes place in which the team members come together to share

(9)

their suggestions. During the group session, team members have to take turns presenting their ideas. These ideas will then be discussed by the entire group to evaluate their

significance (Lynch, 2006, p. 69). Hunton and Gold (2010, p. 916) describe the discussion of this type of brainstorming as structured and they state that free interaction is restricted, which results in enough time and opportunity for every person to present their ideas. Because audit team members have to present their ideas, participation of every member is forced during the session (Hunton & Gold, 2010, 916).

Some research has been carried out to investigate the effectiveness of both group and nominal fraud brainstorming sessions. In general, findings suggest two things. Group fraud brainstorming sessions identify more quality fraud ideas, while nominal fraud

brainstorming sessions identify a higher number of fraud ideas. One study which examined and compared both types of sessions was conducted by Carpenter, Reimers and Fretwell (2011, p. 211). Their results were similar to those of other studies. They find that group fraud brainstorming sessions generate more quality fraud ideas because of increased group interaction. According to the authors, nominal fraud brainstorming sessions produce more ideas about fraud due to a greater emphasis on brainstorming itself (Carpenter, Reimers & Fretwell, 2011, p. 222).

2.5 Face-to-face and electronic fraud brainstorming sessions

Another decision that the auditor has to make is whether or not to brainstorm about fraud in an electronic way. For this decision Lynch, Murthy and Engle (2009, p. 1210) distinguish two different methods: face-to-face fraud brainstorming sessions and electronic fraud

brainstorming sessions.

Face-to-face fraud brainstorming sessions are meetings between the audit team members which happen physically (Lynch, Murthy & Engle, 2009, p. 1210). Every member of the audit team has to be present at the session and communication between members happens in person. No electronic communication technology is used during the session (Lynch, 2006, p. 69). In a face-to-face brainstorming session, group members must wait for their turn to speak and social norms order people to listen to the person who is speaking and not to interrupt him. This means that people cannot share ideas about fraud simultaneously (Lynch, Murthy & Engle, 2009, p. 1210). Because members communicate personally without the use of technology, nonverbal communication is also present during the sessions. This can build trust and collegiality between audit team members (Beasley & Jenkins, 2003, p.37).

Lynch (2006, p. 69) describes electronic fraud brainstorming sessions as fraud brainstorming sessions which involve the use of technology. Audit team members submit their ideas about fraud to other group members by, for instance, group support systems. This

(10)

type of brainstorming session rules out verbal communication between members (Landis, Jerris & Braswell, 2008, p. 71). As a result, people can share their ideas simultaneously because they do not have to wait for another person to finish speaking. Therefore, this type of brainstorming session usually takes less time (Wood & Pickerd, 2011, p. 65). Beasley and Jenkins (2003, p. 37) state that group size is not a limiting factor for this brainstorming

session type. Large groups can communicate with each other without difficulty, since they do not have to be in the same location for the session. Because group members do not meet in person and nonverbal communication is absent, some suggestions about fraud might not be understood properly and misunderstandings between audit team members might arise (Beasley & Jenkins, 2003, p. 37).

Some scientists have conducted research about the difference in effectiveness

between face-to-face and electronic fraud brainstorming sessions. Generally, studies indicate that electronic fraud brainstorming sessions outperform face-to-face fraud brainstorming sessions. For example, research by Smith, Murthy and Engle (2012, p. 334) shows that electronic fraud brainstorming sessions are more effective than face-to-face sessions. The primary reason they provide is that electronic fraud brainstorming sessions have a greater degree of task focus. This means that audit team members who brainstorm electronically are more focused on the main reason for the session, which is to generate ideas about possible fraud risks (Smith, Murthy & Engle, 2012, p. 351).

2.6 Accountable and anonymous fraud brainstorming sessions

The last choice which an auditor faces is the possible use of anonymity during a fraud brainstorming session. Lynch (2006, pp. 69-70) describes two different options for this decision: accountable fraud brainstorming sessions and anonymous fraud brainstorming sessions.

Accountable fraud brainstorming sessions are brainstorming sessions in which audit team members are responsible for the ideas which they generated (Lynch, 2006, pp. 69-70). For every idea it is clear by which group member it was generated and this means that they can be held accountable for those ideas. Connolly, Jessup and Valacich (1990, p. 692) state that this can cause participants to be cautious when submitting ideas about possible fraud. Audit team members might fear negative responses from their fellow team members. Accountability also causes required participation during a fraud brainstorming session. Because every group member’s contribution can be identified, everyone is forced to provide ideas about possible fraud (Pissarra & Jesuino, 2005, p. 278).

Anonymous fraud brainstorming sessions are described by Jessup, Connolly and Galegher (1990, p. 314) as brainstorming sessions in which contributions of individual

(11)

participants cannot be identified. This means that during such a session audit team members generate ideas about fraud, but none of these ideas can be connected to the individual who came up with it. Because of this, group members are less afraid to participate during the session. Inexperienced or introvert audit team members will not hesitate to suggest ideas and non-conventional or less common ideas can also be proposed more easily (DeZoort, Harrison & Taylor, 2006, p. 385). Anonymity can also result in less participation of audit team members. Because an individual’s suggestions cannot be identified, it cannot be noticed if this person participated during the session. Therefore, members may not be motivated to contribute to the brainstorming session (Chen, Zhang & Latimer, 2010, p. 2).

The difference in effectiveness between both types of fraud brainstorming has been investigated by several studies. The overall conclusion of these studies is that the use of anonymity has a positive effect on the number of ideas produced, but that the quality of ideas is equal for both accountable and anonymous fraud brainstorming sessions. This result is supported by Connolly, Jessup and Valacich (1990, p. 699). Their study indicates that anonymous brainstorming causes people to provide more suggestions about possible fraud, since anonymity takes away their fear of rejection by other audit team members. They also find that the quality of ideas does not differ between accountable and anonymous

brainstorming. This is because participants are usually confident about sharing high-quality ideas (Connolly, Jessup & Valacich, 1990, p. 699).

3. Analysis

In this section an analysis of the existing literature will be performed in order to answer the research question. For each decision that the auditor has to make, the different options from which he can choose will be compared to each other. This should result in a clear answer to the research question about the most effective way to carry out a fraud brainstorming

session. As was mentioned in the previous section, a fraud brainstorming session is effective when many ideas about possible fraud have been generated and when generated ideas are of high quality. Therefore, all three comparisons are split into two different components: quality of ideas and quantity of ideas. The first subsection will compare group and nominal fraud brainstorming sessions. In the next subsection the possible use of electronic

brainstorming will be discussed. Finally, the last subsection will show whether participants of fraud brainstorming sessions should be accountable for their ideas or if they should remain anonymous.

(12)

3.1 Group and nominal fraud brainstorming sessions compared

In this subsection an analysis will be made about the difference between effectiveness of group fraud brainstorming sessions and nominal fraud brainstorming sessions. This will be done using several studies which have researched the effectiveness of both types of fraud brainstorming sessions. First, there will be a comparison of the quality of ideas produced during both types of sessions. After that, the difference in quantity of ideas generated will be compared.

3.1.1 Comparison in terms of quality of ideas

Several studies have been carried out to examine the difference between the quality of ideas produced during group and nominal fraud brainstorming sessions. The studies and their outcomes will be discussed in order to find out which method produces the best ideas in terms of quality.

Carpenter (2007, pp. 1132) conducted a study to examine the differences in quality of ideas between group and nominal fraud brainstorming sessions. Results from this

experiment suggested that audit teams generate more quality fraud ideas during a fraud brainstorming session than nominal groups. The study shows that audit teams are more capable of coming up with quality ideas than all the individual members of the audit team combined (Carpenter, 2007, p. 1135). This is because the audit team has the ability to build on fraud ideas suggested by individuals during a session. This makes sure that more quality is added to the proposed ideas. Carpenter (2007, pp. 1136) also states that group

brainstorming leads to the generation of new quality fraud ideas during the session which were not previously identified by individual participants. This is due to the fact that groups benefit from stimulation and synergy. This study examined fraud brainstorm sessions

specifically and provides a clear result regarding the comparison of both methods in terms of quality of ideas. It clearly shows that group fraud brainstorming outperforms nominal fraud brainstorming sessions. This is in line with general expectations about the outcome of this topic.

A study performed by Carpenter, Reimers and Fretwell (2011, p. 219) provides the same result. The authors showed that group fraud brainstorming provides significantly more quality fraud ideas than nominal fraud brainstorming. The study also tried to find out why this is the case. By having auditors brainstorm individually, the researchers were able to separate the effects of brainstorming and group interaction. Results showed that group interaction is the main cause for group fraud brainstorming sessions to produce ideas of higher quality than nominal fraud brainstorming sessions (Carpenter, Reimers & Fretwell, 2011, p. 219).

(13)

Like the study mentioned earlier, fraud brainstorming sessions were examined specifically. This means that the results of this study are significant to the research question discussed in this paper. The outcome of this study is also similar to the results which were expected in advance.

Alon and Dwyer (2010, p. 240) have conducted research which produces a similar result. This study, however, is not entirely relevant to the issue dealt with in this paper. Although the research involves fraud brainstorming sessions, the purpose of the study is to find out what the effect is of the use of decision aids on group brainstorming performance. A decision aid is a procedure which improves the generation of ideas during a brainstorming session (Alon & Dwyer, 2010, p. 242). The authors tested if brainstorming groups with the decision aid would generate more quality fraud ideas than individuals with a decision aid. They found that group fraud brainstorming sessions were able to produce more quality fraud ideas than nominal fraud brainstorming sessions when the decision aid was provided. They therefore conclude that group fraud brainstorming is a beneficial addition to the fraud risk assessment process (Alon & Dwyer, 2010, p. 249). This result is similar to the outcomes of the studies discussed earlier. However, because this study is not completely relevant to the subject of this thesis, it is not as significant as the other studies. Still, it provides a clear outcome which contributes to the result which is commonly expected.

Contrary to the results of the previously mentioned studies, Hunton and Gold (2010, p. 930) find that nominal fraud brainstorming sessions result in ideas which are of higher quality than ideas generated during group fraud brainstorming sessions. While the study also looks specifically at fraud brainstorming, there are still some limitations which make the result questionable. The studies mentioned above based their results on direct measures of quality of fraud ideas. Hunton and Gold (2010, p. 931), however, assessed a second-order effect of quality: audit plan adjustments. Because of this, the results are less accurate. This makes the research less reliable and this decreases the significance of this study.

Dornburg, Stevens, Hendrickson and Davidson (2009, p. 524) also conducted a study which provided different results than expected. They found that idea quality was higher in nominal brainstorming sessions than in group brainstorming sessions. This outcome is not in line with the previously mentioned results and also differs from expectations. There are several reasons to explain this. The first reason is that this experiment is based on regular brainstorming sessions (Dornburg, Stevens, Hendrickson & Davidson, 2009, p. 520). Even though this might be fairly similar to fraud brainstorming sessions, results among both might still vary because of this difference. Furthermore, this study involves electronic brainstorming sessions. Even though the results obtained might be applicable to the use of electronic brainstorming, this does not mean that the same results would have been produced if regular brainstorming was used. Because of these limitations, the significance of this study is

(14)

doubtful. Therefore, the outcome is not as valuable as the results of the studies mentioned earlier.

When comparing the different studies and analysing their results, it is clear that group fraud brainstorming sessions outperform nominal fraud brainstorming sessions in terms of idea quality. The research which was mentioned first was most relevant to the issue discussed in this paper and showed the most significant results. These studies provided clear evidence to show that fraud brainstorming in groups generates the highest idea quality. Other research showed different results. However, these studies contained several

limitations which decrease their significance. Because of this, their evidence is not very reliable and the overall outcome of this analysis is not changed by those results.

3.1.2 Comparison in terms of quantity of ideas

The difference between group fraud brainstorming sessions and nominal fraud brainstorming sessions in terms of quantity has been researched in the fields of both auditing and

psychology. Important studies which have investigated this subject will be analysed in order to find out which type of fraud brainstorming session produces most ideas.

A study carried out by Carpenter (2007, p. 1119) compared both session types to discover which one generated most ideas about possible fraud. The outcome was that

nominal fraud brainstorming sessions produced an overall higher number of ideas than group fraud brainstorming sessions. This result is similar to the general expectations. According to Carpenter (2007, pp. 1134-1135), nominal fraud brainstorming enhances the amount of ideas because individuals can focus on brainstorming itself. They are not distracted by other audit team members’ suggestions, which makes sure that they can concentrate on their own task. This study clearly shows the result which was expected beforehand. The outcome is very significant to the issue discussed in this paper, since it examines fraud brainstorming sessions specifically.

Research performed by Carpenter, Reimers and Fretwell (2011, p. 211) shows the same result. They find that nominal fraud brainstorming sessions generate more fraud risk ideas than group fraud brainstorming sessions. During the study, nominal fraud

brainstorming sessions produced a higher amount of ideas due to several reasons. Because nominal sessions allow audit team members to brainstorm individually, everyone is forced to participate. Group members cannot simply rely on other members’ ideas and have to come up with suggestions themselves (Carpenter, Reimers & Fretwell, 2011, pp. 217-218). This is the primary reason for nominal sessions to generate more ideas than group sessions. This study also focuses particularly on fraud brainstorming sessions. Therefore, the results are

(15)

very important for the overall analysis of both fraud brainstorming methods. A clear result is provided and this is also in line with the general anticipated outcome.

Rietzschel, Nijstad and Stroebe (2006, p. 244) performed a study which also looked at the amount of ideas generated during both nominal brainstorming sessions and group brainstorming sessions. This study differs from the research mentioned earlier because it does not involve fraud brainstorming sessions. Despite this difference, the outcome of this study is similar: nominal brainstorming sessions outperform group brainstorming sessions when it comes to the amount of generated ideas. The authors report that the amount of mandatory tasks during a brainstorming session has an effect on the amount of suggestions proposed by group members. If a brainstorming session requires fewer tasks to be

performed, then more ideas will be produced (Rietzschel, Nijstad & Stroebe, 2006, p. 249). During a nominal brainstorming session there is only one task which a participant has to do: generate ideas about a specific topic. A group brainstorming session also requires

participants to discuss proposed ideas together and this session type is therefore less efficient in idea production. This study provides more evidence to support the general

expectation of nominal fraud brainstorming sessions to outperform group fraud brainstorming sessions. But, as mentioned before, this study examined regular brainstorming sessions. This reduces the significance of this study for the issue discussed in this paper. Still, the outcome clearly shows that nominal sessions produce more ideas than group sessions and this can possibly be applied to fraud brainstorming sessions.

Similar research was conducted by Putman and Paulus (2009, p. 23). Their study also involved regular brainstorming sessions and they found the same result: nominal

brainstorming sessions generate more ideas than group brainstorming sessions. The authors report that their research found that the number of words used to express ideas is an

important factor that influences the amount of ideas generated. Their study showed that group brainstorming sessions produced half the amount of ideas that nominal brainstorming sessions created. This was caused by the fact that group sessions used twice as many words to describe an idea (Putman & Paulus, 2009, p. 34). Long descriptions take up more time and this reduces the efficiency. Those longer descriptions in group brainstorming are used because group sessions require more tasks to be performed. Also, idea generation in groups occurs in a more conversational style. This conversational style may require more words and this causes less ideas to be generated in a specific period of time (Putman & Paulus, 2009, p. 34). This study produces the same outcome as the ones mentioned earlier. However, it was carried out examining regular brainstorming sessions which do not look specifically at fraud. Because of this, the outcome of this study is less important and this should be taken into consideration when a final conclusion on this issue is drawn.

(16)

Dennis and Valacich (1993, p. 531) produced a contrasting result from their research. They state that group brainstorming sessions commonly generate more ideasthan nominal brainstorming sessions. They discovered this using electronic brainstorming sessions which were not aimed at detecting fraud. Their experiment showed that when there are many participants, group brainstorming outperformed nominal brainstorming. With fewer team members, group brainstorming generated an equal amount of ideas as nominal

brainstorming. The authors attribute these results to the ability of electronic brainstorming to introduce few process losses, while process gains are enabled (Dennis & Valacich, 1993, p. 531). For instance, participants have no problems sharing ideas at the same time and free riding is impossible. Also, synergy is created because of the group process. There are some limitations, however, which make the outcome of this study less significant. It did not focus particularly on fraud brainstorming sessions, instead it looks at brainstorming sessions in general. In addition, the experiment which was conducted was based on electronic brainstorming. The result might have been different if face-to-face brainstorming was

investigated. These limitations decrease the importance of the outcome and also the study’s value for this analysis.

Several studies have investigated the difference between group and nominal fraud brainstorming sessions in terms of the number of ideas produced. The overall conclusion seems very clear: nominal fraud brainstorming sessions produce a greater amount of ideas than group fraud brainstorming sessions. The research which was mentioned first was most important to the analysis, since these studies specifically investigated fraud brainstorming. These studies provided a clear result and this result was in line with expectations. Research about regular brainstorming sessions provided similar results. Only one study provided a different outcome. Although it showed that group brainstorming sessions produced more ideas than nominal brainstorming sessions, the study had some limitations. Because of this, the conclusion remains unchanged.

3.2 Face-to-face and electronic fraud brainstorming sessions compared

In this subsection the effectiveness of face-to-face and electronic fraud brainstorming

sessions will be analysed and compared to each other. Research which has been conducted about both fraud brainstorming session types will be used for this analysis. The differences between the types in terms of quality will be compared first. A comparison of both types in terms of quantity will be performed in the second part.

(17)

3.2.1 Comparison in terms of quality of ideas

The difference between idea quality produced by face-to-face and electronic fraud

brainstorming sessions has been examined by several researchers. The studies which they conducted and their results will be discussed and analysed in order to find out which of these fraud brainstorming methods generates the best ideas in terms of quality.

Research carried out by Lynch, Murthy and Engle (2009, p. 1228) shows that audit teams participating in electronic fraud brainstorming sessions identify more high-quality fraud risks compared to the use of face-to-face fraud brainstorming sessions. They state that electronic brainstorming sessions are more efficient because participants can share ideas at the same time without having to wait for one another or being interrupted. Also, audit team members do not have to be in the same place in order to brainstorm about possible fraud. In addition, the authors mention that the use of technology ensures that all ideas mentioned are recorded and this makes sure that no ideas are forgotten or lost (Lynch, Murthy & Engle, 2009, p. 1228). This study particularly examines fraud brainstorming sessions. Therefore, the outcome is very relevant for the analysis performed in this subsection. The outcome is clear: electronic fraud brainstorming sessions generate more ideas of high quality than face-to-face fraud brainstorming sessions. This is similar to what was expected in advance.

A study carried out by Smith, Murthy and Engle (2012, p. 345) provides the same outcome. They also found that the number of identified high-quality fraud risks was higher in electronic fraud brainstorming sessions than in face-to-face fraud brainstorming sessions. According to the authors, the main reason for this is the higher degree of task focus of participants during electronic brainstorming sessions. This means that audit teams which are communicating electronically are more focused on the most important task of the

brainstorming session, which is to come up with suggestions about possible fraud in an organisation (Smith, Murthy & Engle, 2012, p. 351). The authors state that this greater degree on task focus is attributable to the depersonalised environment of an electronic fraud brainstorming session. There is low social presence and communication occurs via electronic technology. This makes it easier for team members to suggest their ideas about possible fraud because there is less social pressure (Smith, Murthy & Engle, 2012, p. 351). Since this study specifically looks at fraud brainstorming sessions, the result is very relevant for the analysis about the difference between electronic and face-to-face fraud brainstorming

sessions. The study provides a clear result: electronic fraud brainstorming sessions generate more high-quality ideas than face-to-face fraud brainstorming sessions. This result is similar to the result of the study which was previously mentioned and it is also in line with

(18)

DeRosa, Smith and Hantula (2007, p. 1569) also researched the difference between electronic and face-to-face brainstorming sessions in terms of quality. Even though the study did not focus on fraud brainstorming sessions, the outcome is similar to the results of the studies mentioned before. The authors found that electronic brainstorming produced better quality ideas than face-to-face brainstorming, which is in line with general expectations. Electronic brainstorming allowed participants to be more focused on their task because they were not distracted by other team members (DeRosa, Smith & Hantula, 2007, p. 1569). This study provides more evidence that confirms the anticipated outcome. However, as

mentioned before, it does not particularly look at fraud brainstorming sessions. Because of this, the result of this study is not as relevant as the outcomes mentioned earlier in this paper.

A similar study was carried out by Gallupe, Dennis, Cooper, Valacich, Bastianutti and Nunamaker (1992, p. 357). They also examined idea quality for both electronic and face-to-face brainstorming. The outcome of the study shows that electronic brainstorming sessions generated more high-quality ideas than face-to-face brainstorming sessions. The authors state that electronic brainstorming reduces production blocking. Production blocking happens when many team members participate in the same session. Only one person can speak at the same time and this makes sure that other participants cannot share their ideas.

Electronic brainstorming solves this problem because suggestions can be made

simultaneously (Gallupe, Dennis, Cooper, Valacich, Bastianutti & Nunamaker, 1992, p. 364). Also, evaluation apprehension is reduced during electronic brainstorming sessions. This refers to the fact that people can be afraid to share their ideas when other people are present during a session. They may fear criticism from their fellow team members. When a

brainstorming session occurs electronically, less people are in the same place and this means that participants are more comfortable sharing their ideas (Gallupe, Dennis, Cooper, Valacich, Bastianutti & Nunamaker, 1992, p. 364). This study reports that electronic

brainstorming sessions produce more ideas of high quality than face-to-face brainstorming sessions. Although this result is similar to expectations, the study examined regular

brainstorming sessions which are not aimed at fraud detection. This reduces the significance of the study for this analysis.

Research performed by Barki and Pinsonneault (2001, p. 158) showed that electronic brainstorming sessions do not outperform face-to-face brainstorming sessions in terms of quality. The authors state that face-to-face brainstorming generated ideas of which the quality was at least as good as the idea quality of electronic brainstorming. However, there are some aspects that limit the importance of this study. The study examined regular

brainstorming instead of fraud brainstorming. Therefore, this outcome may not be applicable to fraud brainstorming sessions. In addition, three conditions which were thought to improve

(19)

electronic brainstorming were manipulated in the experiment (Barki & Pinsonneault, 2001, p. 160). This may have changed the result of the study. Because of these limitations, this study is not as significant as the studies mentioned before and the outcome is not very relevant for the analysis of the difference in idea quality.

The difference between electronic fraud brainstorming sessions and face-to-face fraud brainstorming sessions in term of quality has been the subject of several studies. Research which was mentioned first is most important to this analysis because fraud brainstorming is discussed specifically. These studies clearly showed that electronic fraud brainstorming sessions generated most high-quality ideas, which is similar to the

expectations. Research which looked at regular brainstorming sessions is less significant to the analysis. The studies mentioned first produced the same result as the studies about fraud brainstorming. The last one, however, provided an opposing outcome. Because of the

limitations of this study, the overall conclusion of the analysis is not affected.

3.2.2 Comparison in terms of quantity of ideas

Face-to-face and electronic fraud brainstorming sessions have been compared to each other by several studies in order to find out which method produces most fraud ideas. These studies and their outcomes will be discussed and analysed. This will show whether face-to-face or electronic fraud brainstorming sessions produce most suggestions about possible fraud.

Kerr and Murthy (2009, p. 253) conducted a study in which they compared face-to-face fraud brainstorming sessions to electronic fraud brainstorming sessions in terms of idea quantity. This comparison led to the conclusion that electronic fraud brainstorming sessions generate more fraud ideas than face-to-face fraud brainstorming sessions. This result is similar to the outcome which was anticipated beforehand. Because of this result, the authors recommend the use of electronic fraud brainstorming to auditors who have to carry out a fraud brainstorming session during a financial statement audit (Kerr & Murthy, 2009, p. 258). Since fraud brainstorming sessions are examined specifically, the outcome of this study is very significant for the analysis. There are no limitations to this study, which makes the result relevant. The outcome is in line with expectations and this confirms that electronic fraud brainstorming sessions produce most ideas about potential fraud risks.

A similar study provided the same result. Smith, Murthy and Engle (2012, p. 348) also compared face-to-face fraud brainstorming sessions and electronic fraud brainstorming sessions in order to find out which method produces most fraud ideas. They found that most suggestions about possible fraud were generated during electronic fraud brainstorming sessions. According to the authors, the reason for this is that electronic brainstorming uses

(20)

less procedural commentary. This means that during an electronic fraud brainstorming session there is less communication about the brainstorming activity itself. Participants are all divided over several locations and communicate via electronic technology, which causes them to communicate less among each other about the task at hand. Therefore, audit team members have more time and opportunities to come up with suggestions about fraud. This leads to a higher number of fraud risk ideas (Smith, Murthy & Engle, 2012, p. 348). The outcome of this research is in line with expectations and it is important to the analysis made in this section. This is because the study focused particularly on fraud brainstorming

sessions and this study contains no significant limitations.

Valacich, Dennis and Collonny (1994, p. 463) conducted research which looked at the number of ideas generated during face-to-face and electronic brainstorming sessions. They found that brainstorming groups using electronic technology produced more ideas than groups which brainstormed face-to-face. The authors report that the main reason for this is the elimination of production blocking during electronic brainstorming sessions. Face-to-face brainstorming does not allow multiple participants to suggest ideas at the same time.

Electronic brainstorming is not affected by this problem because people can propose suggestions simultaneously. This makes the session more effective and it results in more ideas generated by participants (Valacich, Dennis & Collonny, 1994, p. 463). This study also shows that electronic brainstorming outperforms face-to-face brainstorming, which is the outcome which was anticipated beforehand. However, the experiment looked at regular brainstorming sessions instead of fraud brainstorming sessions. This reduces the significance of the outcome for this analysis.

Another study also examined face-to-face and electronic brainstorming sessions in terms of idea quantity and this research produced the same outcome. Gallupe, Bastianutti and Cooper (1991, p. 140) compared both brainstorming methods and found that electronic brainstorming produced most ideas. According to the researchers, there are multiple

explanations for this result. They state that communication via electronic technology is faster than writing or talking and visual information is processed quicker than verbal information (Gallupe, Bastianutti & Cooper, 1991, pp. 140-141). This allows more ideas to be generated because there is more time to focus on brainstorming itself. In addition, the technology used during electronic brainstorming sessions contains a built-in memory. Because of this, participants can retrieve all proposed ideas. This can stimulate their own thoughts and it makes sure that no similar suggestions are made twice (Gallupe, Bastianutti & Cooper, 1991, p. 142). This study found the same result as the research which was mentioned before and this is in line with expectations. However, there was no focus on fraud brainstorming

sessions. Because of this, the study is less relevant to the analysis about face-to-face and electronic fraud brainstorming sessions.

(21)

A contrasting outcome resulted from research conducted by Pinsonneault, Barki, Gallupe and Hoppen (1999, p. 110). They also investigated regular brainstormingin order to find out whether face-to-face or electronic brainstorming sessions produced more ideas. Their study showed that groups which brainstormed face-to-face generated more

suggestions than groups which brainstormed electronically. The authors state that the use of electronic technology during brainstorming sessions creates an illusion of productivity. This brainstorming approach is used frequently because participants think that it is more effective than face-to-face brainstorming, while the outcome of this study suggests it is not

(Pinsonneault, Barki, Gallupe & Hoppen, 1999, p. 110). This study did not examine fraud brainstorming sessions in particular, it looked at regular brainstorming sessions instead. As a result, the outcome might not apply to fraud brainstorming sessions. This reduces the

significance of the result for the analysis made in this section.

Multiple studies have examined the difference between face-to-face fraud

brainstorming sessions and electronic fraud brainstorming sessions in terms of idea quantity. In advance of this analysis it was expected that electronic fraud brainstorming sessions would produce most suggestions about possible fraud in an organisation. The studies which examined fraud brainstorming sessions in particular confirmed this: the research clearly showed that electronic fraud brainstorming sessions provided more ideas than face-to-face fraud brainstorming sessions. Research which examined regular brainstorming sessions were less relevant to this analysis because the results provided by these studies might not be applicable to fraud brainstorming sessions. Most of this research also showed that electronic brainstorming generated more ideas than face-to-face brainstorming. One study, however, showed an opposing outcome. Because all other research clearly showed that electronic brainstorming produces more ideas than face-to-face brainstorming, this study is an exception which does not influence the overall outcome of this analysis.

3.3 Accountable and anonymous fraud brainstorming sessions compared

The effectiveness of accountable and anonymous fraud brainstorming sessions will be analysed in this subsection. These different fraud brainstorming types will be compared in order to find out which type is most effective. This will be done using studies which have examined both methods. First, there will be a comparison in terms of quality. After that, the difference in terms of quantity will be compared.

(22)

3.3.1 Comparison in terms of quality of ideas

The difference between accountable and anonymous fraud brainstorming sessions in terms of idea quality has been examined and compared by several studies. The research which has been conducted and the results which were found will be discussed and analysed. This will show whether accountable or anonymous fraud brainstorming sessions generate the best ideas about possible fraud in an organisation.

Cockrell and Stone (2011, p. 238) compared accountable and anonymous fraud brainstorming sessions in order to find out which one produces the best ideas about possible fraud. Their research showed that fraud brainstorming groups in which participants were held accountable for their suggestions resulted in identification of higher quality fraud risks. They state that brainstorming anonymously prevents the creation of high-quality ideas. This is because anonymous brainstorming does not allow audit team discussions, since this would take away the participants’ anonymity (Cockrell & Stone, 2011, p. 238). However,

discussions among audit team members are necessary in order to add quality to proposed ideas. During accountable fraud brainstorming sessions it is possible to have group

discussions and, according to the authors, this is the main reason why it creates ideas of high quality (Cockrell & Stone, 2011, p. 238). This study is very relevant to the analysis made in the subsection because it specifically looks at fraud brainstorming sessions. The expected result beforehand was that there would not be a difference between accountable and

anonymous fraud brainstorming sessions in terms of idea quality. This study clearly showed that accountable fraud brainstorming sessions produced better suggestions about possible fraud, so this differs from the expectations.

Research conducted by Lynch (2004, p. 91) also investigated the difference between accountable and anonymous fraud brainstorming sessions. The outcome of this study was that accountable fraud brainstorming results in ideas of higher quality than anonymous fraud brainstorming. The author states that the main reason for this result is that accountability leads to more pressure on audit team members to generate high-quality ideas. Participants are evaluated based on their idea production and this causes them to put in more effort in order to make a good impression on their superiors (Lynch, 2004, p. 91). During anonymous fraud brainstorming sessions it is possible for audit team members to put in less effort because their superiors will not find out who came up with which idea. Participants can simply benefit from their team members’ idea production and this causes less ideas to be generated (Lynch, 2004, p. 91). Because this study particularly examined fraud

brainstorming sessions, it is very relevant for the comparison of both brainstorming methods in terms of idea quality. This study also showed that more high-quality ideas are generated

(23)

when participants are held accountable for their suggestions. This result is similar to the result of the study mentioned before, but it is different from the anticipated outcome.

Hunton and Gold (2010, p. 932) examined the different ways in which a fraud

brainstorming session can be carried out. They looked at the results of the different types of sessions in order to find out which method produces the best ideas in terms of quality. They found that anonymity of audit teams during fraud brainstorming sessions results in a lower number of high-quality ideas generated (Hunton & Gold, 2010, p. 932). This means that teams in which participants are accountable for their suggestions about fraud produce ideas of higher quality than groups which brainstorm anonymously. This result is similar to the results of the studies mentioned before, but it is not in line with the anticipated outcome. Because this study specifically looked at fraud brainstorming sessions, the outcome is very useful for the analysis in this subsection.

A study performed by Connolly, Jessup and Valacich (1990, p. 698) examined regular brainstorming sessions in order to find out what the effect of anonymity is on the quality of ideas produced during these sessions. They found that accountable brainstorming sessions produce ideas of which the quality is slightly better than ideas generated during anonymous brainstorming sessions. The authors report that there was no noticeable difference in

average idea rarity between accountable and anonymous brainstorming sessions. However, accountable brainstorming groups scored higher than anonymous brainstorming groups on the rarity of their single rarest idea. Therefore, it is preferable to make participants

accountable for their suggestions (Connolly, Jessup & Valacich, 1990, p. 698). The result of this study is quite similar to the results which were mentioned before. However, this study looked at regular brainstorming sessions. This reduces its significance for the comparison of accountable and anonymous fraud brainstorming sessions.

An opposing result was found by Pissarra and Jesuino (2005, p. 285). They researched whether quality of ideas generated during regular brainstorming sessions benefited from anonymity of the participants. According to the authors, idea quality was higher when brainstorming happened anonymously. They state that anonymity reduces participants’ fear of possible disagreement of other group members, allowing them to propose more high-quality ideas (Pissarra & Jesuino, 2005, p. 286). This study provides a different outcome than the research mentioned before. This is explained by the fact that it contains multiple limitations. This study did not specifically examine fraud brainstorming sessions, which makes the result less significant for the analysis made in this subsection. In addition, some variables related to anonymity were manipulated (Pissarra & Jesuino, 2005, p. 286). This may have altered the result, which also reduces the importance of this study. Because of these limitations, the result does not influence the overall conclusion of this analysis.

(24)

The difference between accountable and anonymous fraud brainstorming sessions in terms of idea quality has been examined by several studies. The expected result before this analysis was that there would be no difference in idea quality between the two methods. The research which looked at fraud brainstorming sessions provided another result: fraud

brainstorming sessions in which participants were accountable for their suggestions provided ideas of higher quality than anonymous fraud brainstorming sessions. Studies which

examined regular brainstorming sessions were less important to this analysis because the outcomes might not apply to fraud brainstorming sessions. Most of this research also

showed that the use of accountability resulted in higher quality ideas. One study provided the opposite result. However, it contained several limitations. Because of this, the overall

conclusion of this analysis remains unchanged.

3.3.2 Comparison in terms of quantity of ideas

Multiple studies have examined the number of ideas produced during accountable and anonymous fraud brainstorming sessions. The research which has been carried out and the outcomes which were found will be discussed and analysed in order to find out which fraud brainstorming method produces most ideas about possible fraud in an organisation.

A study conducted by Cockrell and Stone (2011, p. 238) looked at both accountable and anonymous fraud brainstorming sessions to find out which method produces most fraud ideas. They report that fraud brainstorming sessions in which participants are anonymous generate the least number of suggestions. This means that during accountable fraud brainstorming sessions the group members produce a larger amount of ideas. According to the authors, this is caused by the benefits which are derived from the team discussion. This discussion takes place during accountable fraud brainstorming sessions. It stimulates the audit team members to generate more ideas and it also makes sure no comparable ideas are produced twice (Cockrell & Stone, 2011, p. 238). The only way in which anonymity of participants can possibly benefit fraud brainstorming sessions is when the audit team has to identify frauds which have never occurred before and which are difficult to detect. Anonymity can then take away fear of group members to propose ideas which are less common and this increases idea productivity. These situations rarely take place, so this does not change the conclusion of this research (Cockrell & Stone, 2011, p. 238). This study focused on fraud brainstorming sessions specifically. Therefore, it is very useful for the analysis made in this subsection. The expectation beforehand was that anonymity of participants would increase the number of ideas being produced during fraud brainstorming sessions. This study, however, provides the opposite outcome.

(25)

Lynch (2004, p. 91) performed a similar study which also looked at the number of ideas generated during accountable and anonymous fraud brainstorming sessions. This study provided the same outcome as the research mentioned before. The author states that accountable fraud brainstorming sessions generate the greatest number of fraud ideas. This is due to the fact that accountability induces a performance-related pressure for participants to generate suggestions about possible fraud. Since every group member who came up with a certain idea is identifiable, all audit team members are forced to participate during the fraud brainstorming session. This causes more ideas to be generated during the session (Lynch, 2004, p. 91). This study particularly examined fraud brainstorming sessions. This means that the result is very important for the comparison made in this subsection. The anticipated outcome was that anonymous fraud brainstorming sessions would produce a higher number of ideas than accountable fraud brainstorming sessions. This study shows a contrasting result, similar to the result of the research mentioned earlier.

Another study which found the same result was carried out by DeZoort and Harrison (2007, p. 3). The authors investigated the effects of both accountability and anonymity of participants on idea production during fraud brainstorming sessions. They found that more ideas were generated during accountable fraud brainstorming sessions. The study shows a clear explanation for this outcome. Audit team members who are held accountable for their ideas experience higher levels of fraud detection responsibility. This causes them to brainstorm more fraud-related procedures, which results in more ideas being generated (DeZoort & Harrison, 2007, p. 3). When participants of fraud brainstorming sessions are anonymous, they feel less responsibility to find fraud in an organisation and this causes their idea production to decrease. Because this research specifically looked at fraud brainstorming sessions, the outcome is useful for the analysis of the possible use of anonymity during these sessions. The result of this study is similar to the results mentioned before and this shows that accountable fraud brainstorming sessions generate most ideas about possible fraud in an organisation.

A different result was found by Jessup, Connolly and Galegher (1990, p. 317). Their study focused on regular brainstorming sessions and it looked at the effect of anonymity on the number of ideas which were produced. The outcome showed that there is no difference between accountable and anonymous brainstorming sessions in terms of the number of ideas generated (Jessup, Connolly & Galegher, 1990, p. 317). Before the experiment it was expected that anonymous brainstorming sessions would produce most ideas. This means that the outcome does not match the result which was anticipated in advance. The result of this study also contrasts the outcomes mentioned earlier in this section. This can be

explained by the fact that this experiment investigated regular brainstorming sessions. Because of this, the finding of this study might not be applicable to fraud brainstorming

(26)

sessions. This decreases the significance of the outcome for the analysis made in this subsection.

Mukahi, Chimoto and Ui (1998, p. 366) also examined regular brainstorming sessions in terms of idea quantity and found a similar result. They report that the potential benefits of anonymity depend on the different personality types of the participants. The authors state that introvert group members generate more ideas when brainstorming is anonymous. Anonymity takes away their fear of possible rejection of their ideas by other participants. Extrovert team members, on the other hand, propose more suggestions when they are accountable for their ideas. The reason for this is that they are able to be involved in

discussions with other participants, which stimulates their productivity (Mukahi, Chimoto & Ui, 1998, p. 366). Because of these findings, the authors recommend that the group leader makes a decision about the use of anonymity based on the personalities of the group members. Although this study’s result is similar to the result of the study previously described, it contrasts the expectations and the results from research about fraud

brainstorming sessions. Since regular brainstorming sessions were investigated in this study, the outcome might not apply to fraud brainstorming sessions. This reduces the relevance of this result for the analysis made about the possible use of anonymity in these sessions.

Several studies have examined the effects of accountability and anonymity of participants on the number of ideas produced during fraud brainstorming sessions. The anticipated outcome was that anonymous group members would produce more ideas than participants who are accountable for their suggestions. This outcome was not found by the research which specifically looked at fraud brainstorming sessions. These studies found that accountability of audit team members produced more ideas than anonymity. Research which looked at regular brainstorming sessions showed that there is no difference between

accountable and anonymous brainstorming in terms of idea quantity. These studies,

however, were less important to the analysis because the outcomes might not be applicable to fraud brainstorming sessions. Therefore, the overall conclusion is not affected by these findings.

4. Discussion

In this section a discussion will take place about the findings of this paper. The purpose of this thesis was to find out in what way a fraud brainstorming session should be carried out to be most effective. The results which were found in the analysis of the previous section will be discussed in order to answer the research question. Also, the implications of these results for

(27)

real-world practice will be mentioned. There are some limitations to this thesis, however, which will be pointed out as well. Finally, suggestions for future research will be provided.

First, the analysis compared group and nominal fraud brainstorming sessions. This comparison did not clearly show which method is most effective. Although group fraud brainstorming sessions produced the highest idea quality, nominal fraud brainstorming sessions generated an overall higher number of ideas. Because the quality of ideas and the number of ideas produced are both equally important for fraud brainstorming sessions, none of these techniques is preferred above the other. Therefore, it is up to the auditor to choose whether group or nominal fraud brainstorming will be used.

The next comparison which was made in the analysis was about the decision between face-to-face and electronic fraud brainstorming sessions. It was shown that when electronic technology is used, the ideas which are generated are usually of higher quality. Also, audit teams which brainstorm electronically produce more ideas about possible fraud in an organisation. Since electronic fraud brainstorming sessions outperform face-to-face fraud brainstorming sessions in both quality and quantity, electronic technology should clearly be used during fraud brainstorming sessions.

In the final part of the analysis, accountable and anonymous fraud brainstorming sessions were compared to find out which method is most effective. The clear conclusion was that audit team members who are accountable for their suggestions produce most ideas about fraud and their ideas are of higher quality than those generated by anonymous

participants. This shows that accountable fraud brainstorming sessions are more effective than anonymous fraud brainstorming sessions. Therefore, the auditor has to make sure that participants can always be identified by the fraud risk ideas they have produced.

The analysis was performed with the objective of answering the research question of this paper. The purpose was to find out what the most effective way is to carry out a fraud brainstorming session. The analysis clearly shows how a fraud brainstorming session should be carried out: it should take place either in a nominal or group session, electronic

technology must be used and audit team members should be accountable for the ideas they propose about possible fraud.

The findings of this paper have important implications for real-world practice. Fraud brainstorming sessions are a required procedure for audit teams during a financial statement audit. However, SAS No. 99 does not state how a session should be carried out. The auditor has the ability to make several decisions about the way in which a session will be conducted. The results of this thesis show the most effective way to carry out a fraud brainstorming session and auditors can use this information when they have to lead a session.

This thesis is subject to several limitations. For the analysis, the existing literature about the different fraud brainstorming methods was used. No actual experiments were

Referenties

GERELATEERDE DOCUMENTEN

It states that there will be significant limitations on government efforts to create the desired numbers and types of skilled manpower, for interventionism of

A method for decomposition of interactions is used to identify ‘smaller’ interactions in a top-down analysis, or the ‘smaller’ interactions can be grouped in more complex ones in

This hypothesis states that motivation has an interacting effect on cognitive style and satisfaction, such that brainstorming group consisting of a high average level of

The most effective strategy overall involved face-to-face recruitment through clinics in physician and ultrasound offices with access to a large volume of women in early

According to Ronaan it is very difficult for U.S. multinationals to generate profit in India without using a business group like Mahindra & Mahindra. Like the U.S. firms have

In line with these outcomes, Manzur and colleagues (2011) estimated that the effect of a price promotion on brand loyalty is lower for higher priced national brands compared to

In order to answer hypothesis 2, a higher number of choice sets leads to differences in attribute importance (price becomes more important than brand), and

In total, 51 different interventions (including 13 cost saving interventions) were identified and ranked based on their incremental cost-effectiveness ratio (ICER) and potential